1-IoT Security Project – Bot-IoT Dataset

End-to-End ML Pipeline for IoT Network Attack Detection

Developed ML models to detect and classify malicious network traffic (DDoS, DoS, Reconnaissance) using Bot-IoT dataset.

Ensured real-world reliability by eliminating data leakage and sensitive features (IP, MAC, timestamps).

Applied advanced feature engineering, data transformation, and outlier handling.

Tested XGBoost and Random Forest models achieving 99.71% and 99.68% accuracy respectively.